Full assessment of Sida (Sida hermaphrodita) biomass as a solid fuel

Due to an increased awareness of climate change and limited fossil resources, the demand for alternative energy carriers such as biomass has risen significantly during the past years. This development is supported by the idea of a transition to a bio‐based economy reducing fossil‐based carbon dioxide emissions. Based on this trend, biomass for energy is expected to be used in the EU mainly for heating until the end of the decade. The perennial herbaceous mallow plant Sida hermaphrodita (L.) Rusby (‘Sida’) has high potential as an alternative biomass plant for energy purposes. Different density cultivation scenarios of Sida accounting for 1, 2, or 4 plants per m2 resulted in a total biomass yield of 21, 28, and 34 tons dry matter/ha, respectively, over a 3‐year period under agricultural conditions while the overall investment costs almost doubled from 2 to 4 plants per m2. Subsequently, Sida biomass was used as SI) chips, SII) pellets, and SIII) briquettes for combustion studies at pilot plant scale. Pellets outcompeted chips and briquettes by showing low CO emission of 40 mg/Nm3, good burnout, and low slagging behavior, however, with elevated NOx and SO2 levels. In contrast, combustion of chips and briquettes displayed high CO emissions of >1,300 mg/Nm3, while SO2 values were below 100 mg/Nm3. Contents of HCl in the flue gas ranged between 32 and 52 mg/Nm3 for all Sida fuels tested. High contents of alkaline earth metals such as CaO resulted in high ash melting points of up to 1,450°C. Life cycle assessment results showed the lowest ecological impact for Sida pellets taking all production parameters and environmental categories into consideration, showing further advantages of Sida over other alternative biomasses. Overall, the results indicate the improved applicability of pelletized Sida biomass as a renewable biogenic energy carrier for combustion.


| INTRODUCTION
Due to an increased awareness of both climate change and limited fossil resources, the demand for alternative energy carriers such as plant biomass has risen significantly during the past years. The use of biomass for energy generation is a central component of sustainable societies. In particular, in conjunction with fluctuating renewable energy sources like wind and solar power, storable biomass is a key to a stable and secure energy supply (Szarka, Eichhorn, Kittler, Bezama, & Thrän, 2017).
The usage of biomass as an energy carrier is supported by the idea of a transition to a bio-based economy reducing fossil-based carbon dioxide emissions (Langeveld, Dixon, & Jaworski, 2010). Ambitious goals were set for the European Union (EU) in the Renewable Energy Directive targeting of at least 32% for the share of renewable energy consumed in the EU in 2030, making the EU a global leader in renewable energy to cut carbon dioxide emissions (European Parliament and Council, 2018). A revised Renewable Energy Directive took into account specific objectives, aiming for 'developing the decarbonisation potential of advanced biofuels' and 'develop renewable energy potential in the heating and cooling sector' (Council of the European Union, 2017)-to which biogenic solid fuels provide a substantial contribution. Based on the current trend, biomass for energy is expected to be used mainly for heating in the future (Council of the European Union, 2014; WBA, 2018).
The perennial herbaceous mallow plant Sida hermaphrodita (L.) Rusby, also known as Virginia fanpetals (hereafter referred to as Sida), has been described in the early 1980s (Spooner & Hall, 1983). The awareness for its potential as an alternative biomass plant for energy purposes increased considerably since 2009 (Borkowska & Molas, 2012;Borkowska, Molas, & Kupczyk, 2009;Howaniec & Smoliński, 2011). Among its use as a biogenic energy source and its potential as a feedstock for combustion, biogas, or hydrogen-rich gas production (Howaniec & Smoliński, 2011;Jablonowski et al., 2017), Sida was subject of few research activities related to its cultivation (Kurucz, Antal, Gábor, & Popp, 2014;Nabel, Temperton, Poorter, Lücke, & Jablonowski, 2016), and its use for phytoremediation on heavy-metal-contaminated soils (Kocoń & Jurga, 2017;Kocon & Matyka, 2012;Krzywy-Gawrońska, 2012). With regard to its agricultural cultivation, the optimal planting density of Sida is crucial in terms of biomass yield and sustainable biomass production (von Gehren et al., 2019). Domestically produced, Sida biomass can contribute to national energy security due to its high caloric value and its tolerance for a broad range of soil conditions . In terms of sustainability and ecosystem services, Sida appears to be very promising due to its perennial and flowering nature and its ability to grow on marginal soils (Nabel, Schrey, Poorter, Koller, & Jablonowski, 2017;Nabel et al., 2016). Particularly, the latter aspect is of importance when it comes to land-use and foodversus-fuel conflicts for biomass production, which could be avoided if the worldwide potential of marginal soils is considered (Cai, Zhang, & Wang, 2011;Krasuska, Cadórniga, Tenorio, Testa, & Scordia, 2010;Zhuang, Jiang, Liu, & Huang, 2011). However, the current scientific data available on Sida as an alternative but promising solid fuel are still scarce.
Within this study, we analyzed different Sida density cultivation scenarios under agricultural conditions, plant establishment, harvest, and yield of Sida biomass over a 4-year period in total, that is, including the year of planting, which was not considered for harvest due to low biomass yields. Subsequently, the Sida biomass was investigated as a solid fuel for combustion, applying Sida in three different forms of dried biomass, that is, SI: shredded biomass, SII: pellets, and SIII: briquettes, followed by detailed ash analysis, economic calculations, and an overall life cycle assessment (LCA).
A key challenge is the heterogeneity of biomass reflected by the wide range of combustion properties of different biomass feedstocks, such as their chemical composition (Jenkins, Baxter, Miles, & Miles, 1999). Accordingly, it is important to gain basic information and practice-relevant data on the combustion behavior of biogenic feedstocks to expand the usable biomass potential and to develop solutions that enable the cost-effective exploitation of alternative fuels.
The necessity of a stable long-term combustion of a biogenic fuel is mainly affected by the elemental composition of the biomass and its ash. Besides the calorific value, the resulting flue gas composition and the chemical composition of the ash are generally considered to classify the usability of a solid fuel. According to Vassilev et al., biomass is often highly enriched in manganese, potassium, phosphorus, and chlorine (Vassilev, Baxter, Andersen, & Vassileva, 2010) and these elements can cause severe technical problems such as sintering or erosion due to acid formation during the combustion process (Vassilev, Vassileva, Song, Li, & Feng, 2017). Especially potassium as an alkaline can lead to severe slagging in a silicate-rich matrix due to a massive reduction of the eutectic temperature.
In the context of environmental and health protection, the emissions of harmful gases, such as nitrogen oxides (NO x ), carbon monoxide (CO), and unburned hydrocarbons or fine dust must be reduced to a tolerable minimum. Against this background, emissions legislation and regulations get tighter not only at European level (European Council, 2008, 2015) but also at national and municipal levels.
To evaluate Sida biomass and its usage considering three different dried biomass forms from an ecological perspective, an LCA was conducted. LCA allows the quantification of system-caused environmental impacts using defined impact categories (ISO14040, 2006). Furthermore, the determination of appropriate boundaries of the systems investigated, combined with available and useful real data, enables ecological sustainability benchmarking with comparable systems.

| Experimental cultivation site and set-up
The experimental field for Sida cultivation was located in Titz Sevenich, Germany (100 m a.s.l., 50°58'13.0"N 6°23'31.0"E, using ETRS89). The field soil was classified as a clayey silt, composed of 5.6% sand (0.063-2.0 mm), 79.0% silt (0.002-0.063 mm), and 15.4% clay (<0.002 mm; all determined in accordance with VDLUFA Methodenbuch Band I, 2012, C 2.2.1), with pH 7.0 (CaCl 2 ). The soil contained 41 kg/ha NO 3 , <1 kg/ha NH 4 , 760 mg/kg P 2 O 5 , 460 mg/kg K 2 O, and 100 mg/ kg Mg, and a humus content of approx. 2.3%, all measured in samples from 0 to 30 cm depth. Prior to planting, the field soil was prepared to obtain a fine crumbly seedbed. Root cuttings of Sida for propagation were obtained from own plantations , and were placed approx. 5 cm below ground using a two-rowed semi-automated potato planter in March 2015. Root cuttings were planted in three different densities to evaluate the plant-density-related biomass yield, and the overall plant performance related to stand density: (a) one root cutting per m 2 ; (b) two root cuttings per m 2 ; and (c) four root cuttings per m 2 , planted in four rows each, with a row distance of 0.75 m and a length of 50 m. This resulted in an experimental field totaling approx. 450 m 2 , with approx. 150 m 2 for each planting density. Subsequently, half of the planting rows were compacted by rolling to evaluate if soil compaction had an effect on germination and shooting. Before Sida re-growth, 100 kg/ha of nitrochalk (equaling 27 kg N/ha) was applied in March 2016 to promote Sida growth and thus suppress weeds. In April 2016, the herbicide GoltixGold (2 L/ha, 700 g/L Metamitron [59.7 weight-%]) was applied to the Sida field at BBCH Sida development stage 14 (i.e., four Sida leafs present; Jablonowski et al., 2017) to control undesired weed growth. In the following, weed was removed manually. For this investigation, Sida biomass was used from three consecutive years (2016, 2017, and 2018), that is, from its second until its fourth year after establishment. Biomass from its establishment year 2015 was not considered for this research due to relatively small quantities.

| Biomass sampling and preparation
For the evaluation of biomass yield related to planting density, dried Sida shoots were collected manually from five individual square meters for each planting density in February to March in three consecutive years (2016, 2017, and 2018). The number of shoots and maximum shoot height of each individual plant was recorded. All samples were freshly weighed, individually shredded (<2 cm) and homogenized at place using a Viking GB 460C, and subsequently dried at 60°C to determine their dry weights using five subsamples of approx. 500 g. Values were extrapolated to obtain the mean dry matter biomass yield per hectare.
For processing and combustion, the entire biomass from the field was harvested, shredded, and homogenized operating a Claas Jaguar crop chopper before Sida re-growth. Residual water content of the entire biomass was determined for recalculation purposes of the biomass as mentioned above, accounting for 19% (±4%) for all three biomass harvests. Biomass was stored and shipped in big-bags for further processing and combustion analysis.

| Statistical analysis
The field experiment has a one factorial design with the factor planting density separated into three levels: 1, 2, or 4 Sida plants per m 2 . Five biological replicates harvested completely randomized in three consecutive years were used for each treatment level. Statistical analysis was performed with analysis of variance in R 3.0.3 (The R Foundation for Statistical Computing 2014) using the work package 'Agricolae' (de Mendiburu, 2014).

| Biomass processing
For biomass combustion and ash analysis, the Sida biomass was either SI) shredded, SII) shredded and pressed into pellets or SIII) shredded and pressed into briquettes. Briquettes were produced employing a Mütek MPP280S briquetting press, with an operational capacity of 150 kg/hr and a power requirement of 8 kWh per 150 kg shredded Sida biomass. The obtained Sida briquettes were of solid nature, with a cylindrical shape of approx. 60 mm height, and 80 mm in diameter, with a bulk density of approx. 660 kg/m 3 .
Pilot scale tests were carried out to evaluate the feasibility of pelletizing Sida, employing a flat die press (screwpress GmbH, KKP 200). The stated power of the engine amounts to 7.4 kW with a maximum throughput of 150 kg/hr. Since the available Sida biomass was limited, the press was preheated with rapeseed cake. The loose Sida chips were mixed with water and molasses (BayWa AG) containing 28.13% C, 7.76% H, 1.67% N, and 0.27% S, using a mixing ratio of 98:1.5:0.5.
A batch of approximately 125 kg Sida pellets was produced and the power consumption and the throughput of pellets were determined. Subsequently, a sieve analysis was performed to determine the fine content of the pellets based on DIN EN ISO 17827-1 (ISO 17827-1, 2016). It is noteworthy that in the case of pelletizing, the combustion behavior can be improved by the addition of additives to reduce the slagging tendency of the fuel (Zeng, Pollex, Weller, Lenz, & Nelles, 2018). In the case of pelletizing Sida, no additives were added to increase the ash softening point, as for Sida ash no slagging problems were expected due to its high melting temperature, as reported in an earlier study .

| Combustion and emission analysis
Sida chips and pellets were tested operating a grate furnace (HDG Bavaria Compact C 100) with a nominal heat output of 100 kW. The furnace was equipped with a moving step grate, and the removal of the grate ash and fly ash was realized separately via two screws into separate containers.
Sida briquettes were tested in a wood burning stove (Carborobot furnace, C40 Bio Farmer), combining a rotary grate for bulk materials with a combustion chamber for log wood or briquettes, and had a nominal heat output of 40 kW.
All systems were equipped with additional temperature sensors, volume flowmeters, and pressure sensors and were connected to a superordinate programmable logic controller allowing an automated evaluation of the test data.
In all experiments, the system was heated up with wood pellets first. After the heating process, the respective Sida solid fuel was applied. Subsequently, the controllable variables such as mass flow, primary and secondary air, negative pressure in the combustion chamber, and the settings of the grate were adapted to allow an optimized operation.
The Sida test fuel was weighed in batches to determine the mass flow of the different Sida fuels, that is, chips, pellets, and briquettes. A Multi Component Analyzer (MCA 04 of Dr. Födisch AG) was employed to analyze and monitor the composition of the flue gas continuously. For CO, CO 2 , NO, NO 2 , NH 3 , SO 2 , HCl, and H 2 O, an infrared absorption analyzer, and for O 2 a zirconia sensor was used. In addition, gravimetric dust measurements according to the specifications of VDI 2066 were carried out (VDI, 2006).
Values of other biomasses such as Miscanthus pellets, wood chips, and log wood used for comparison were obtained from own investigations (unpublished data) employing the same furnace, settings, and approaches, or were extracted from literature, as indicated.

| Chemical and mineralogical analyses of Sida fuels and ashes
To characterize the elemental composition of Sida biomass and its ashes, ultimate analyses were applied. The determination of inorganic elements was performed with inductively coupled plasma with optical emission spectroscopy (ICP OES) for all Sida fuels, namely chips, pellets, and briquettes and their respective ashes obtained from combustion in an ashing furnace at 550°C in air for 24 hr to ensure complete conversion. Subsequently, the ash mineralogy was investigated with X-ray diffraction (XRD) using a Bruker device D4 Endeavor with CuKα anode (40 kV, 40 mA) and a Bragg-Brentano geometry at 0.026° steps with 1 s per step.
To investigate the melting behavior of the ashes, an ash fusion test was performed as follows. Cylindrical pellets with a diameter of 5 mm were pressed with 1 kN preparation pressure and heated to 1,430°C with a heating rate of 5 K/ min. For evaluation and according to Pang et al., the temperature-dependent height was normalized by the initial height (Pang, Hewakandamby, Wu, & Lester, 2013).

| Life cycle assessment analysis
The ecological evaluation by LCA was performed with the software GaBi 9.2 in line with ISO14040 (ISO14040, 2006) and ISO14044 (ISO14044, 2006). The necessary process data were project-generated quantitative results for the main processes and datasets from ecoinvent 3.5 and GaBi 9.2 for the so-called background data (GaBi LCA Databases, 2020; The ecoinvent Database, 2020). Three different cradle-togate process chains for dry Sida biomass pathways were assessed. Intermediate products, and thus technology-determining characteristics of the process chains, are chips in process chain SI, pellets in process chain SII, and briquettes in process chain SIII ( Figure 1). Furthermore, a comparison with alternative biomass heat production was done using existing datasets from the ecoinvent database, that is, wood pellets and logs. The functional unit (FU) for the process chains of the production and use of Sida chips, briquettes, pellets, and comparable processes was defined as the production of 100 GJ thermal energy.
Considered process chains were divided into four stages, namely biomass production (cultivation; harvesting), production/processing (manufacturing chips, pellets, or briquettes, incl. internal logistics), storage and shipping (incl. external logistics), and combustion (customer storage; small scale furnace). Included process modules are presented in Figure 1. As a superordinate setting, a total field size of 4 ha was assumed, whereby applied parameters were generated by the use of smaller test sites. The design and dimensioning of single biomass production process modules are related to this area size by a harvesting equivalent (kg/ha or m 3 /ha). The biomass production stage includes the field preparation, plant protection, and harvesting activities. Calculations of fertilizer-distribution-related emissions (heavy metals, ammonia, etc.) were done as described in Nemecek and Schnetzer (2011). Production/processing of dry biomass describes logistics, storage of raw material, and the necessary machinery for the transformation of biomass. Machines for pelletizing and briquetting were supposed to be provided by external services. Biomass drying was not taken into account since Sida biomass is harvested in naturally dry state. In the storage and shipping stage internal logistics (e.g., conveyor), product storage and suitable technologies for the delivery to the combustion site were considered. The combustion stage includes product storage on site and the technical facility for combustion, in which the furnaces differ by performance classes. Proportionated impacts from manufacturing machines and buildings were included. Biomass losses along the whole process chain were considered by calculations in line with methods described in Seiffert (2010). To ensure the general consistency of Sida process chains with comparable process chains, the disposal of ash was included as an additional process stage.
The datasets of alternative biomass energy carriers and their combustion are based on Bauer (Bauer, 2007). To run the life cycle inventory analysis for the developed process chains, the newest available data with geographical levels for EU and Germany were used. The impact assessment (LCIA) was performed using 16 midpoint impact categories according to the International Reference Life Cycle Data System, out of which five midpoint categories with the largest impact are shown exemplarily. By converting absolute results into normalized values in line with the Product Environmental Footprint (PEF; Pilot 1.09, EU-27 (2010)), it is possible to compare different impacts by person equivalents (PE). PE represent the annual environmental impact of an average European citizen (European Comission, 2016;Hauschild & Wenzel, 2001).
It has to be noted that two variables generally worthy of consideration were omitted due to missing data. The first variable is the quantitative effect of Sida growing to soil organic carbon and the consequences for CO 2 sequestration, which depend on factors such as location, soil type, or methods of cultivation. Especially in relation to LCA studies, this topic is discussed in terms of methods (Albers, Avadí, Benoist, Collet, & Hélias, 2019;Qin, Dunn, Kwon, Mueller, & Wander, 2016). According to Bessou et al. (2019), 'there is still no scientific consensus on the best method to assess the holistic impact of land use and land use change within LCA', which additionally illustrates the issues involved, apart from the missing data. As a second omitted topic, the inclusion of a prolonged cultivation comprising multiple cycles has to be mentioned. The distribution over a variety of cycles following the phase considered here is expected to reduce specific environmental burdens. Based on the data used, only a single cultivation and harvesting phase was considered. A further not discussed topic in the present work is the effect of a yield reduction.

| Plant establishment and biomass productivity
Sida plants grew well leading to a closed canopy after approx. 5 months for all planting densities at BBCH 29 . Plant development during establishment and in the subsequent year was positively supported by herbicide application, manual weed removal, and the application of nitrochalk. Sprouting of the Sida root cuttings accounted for approx. 80% throughout. Missing plants were replanted to guarantee the planting density in the plots. All root cuttings sprouted equally well irrespective of their size and diameter. No differences in root cuttings germination were observed whether the soil was compacted or not.
In the second year after planting, fungal infections were observed on lower Sida stalks starting in BBCH development stage 51, leading to a bending of long stalks. Such infections were more prominent in the highest planting density, that is, 4 plants/m 2 , but were small overall. Stalk bending was also observed in summer and late autumn due to high wind loads.
Biomass yield increased significantly in the second year after establishment, accounting for 10.3, 12.2, and 17.0 t DM/ha at a planting density of 1, 2, or 4 plants per m 2 , respectively ( Figure 2). This reflects a seven-and sixfold biomass increase compared to the first year of harvest in 2016. Except for the planting density of 2 plants per m 2 , the yields decreased moderately in 2018, accounting for 9.6, 13.9, and 14.3 t DM/ha at a planting density of 1, 2, or 4 plants per m 2 , respectively. However, the total produced biomass over all 3 years was highest for the planting density of 4 plants per m 2 , accounting for 34.3 t DM/ha, followed by 28 and 21.4 t DM/ha for the planting density of 2 and 1 plants per m 2 (Figure 2).

| Emissions and combustion characteristics
Wood pellets were used to start the combustion process and heat up the furnace. As soon as the process was stable and the burnout of the gas phase was complete, the three different Sida fuels were injected and emission parameters were evaluated.

| Sida chips
After switching to Sida chips, CO emissions increased immediately due to the lower energy density of the fuel and the resulting sinking temperatures in the combustion chamber from initially about 845°C to about 420°C. Fuel injection was immediately increased stepwise, resulting again in an increased carbon turnover, indicated by an increase in the CO 2 concentration from about 2%-4%. However, the increase in the fuel input was not enough to raise the temperatures in the combustion chamber sufficiently to achieve a stable and complete combustion. Accordingly, the temperatures in the furnace remained at low level and the reaction conditions in the combustion chamber were insufficient for complete conversion of the carbon in the gas phase resulting in high CO emissions, accounting for 1,347.3 mg/Nm 3 (11% O 2 as reference).

| Sida pellets
Sida pellets were of high quality with a shiny surface. This indicated a high degree of polymerization of the lignin in the fuel, and is considered as an indicator of high stability and low abrasion of the pellets. The mean content of fines accounted for 3.8%. After switching to Sida pellets, the system continued to run with the same parameterization as with wood pellets. The uniform CO 2 concentrations and the uniform temperature profile showed a stable reaction. CO emissions remained at a very low level with 40.2 mg/Nm 3 . Concentrations of NO x and SO 2 increased significantly, from 27.2 to 428.6 mg/Nm 3 for NO x , and from 25.9 to 307.2 mg/Nm 3 for SO 2 at 11% O 2 , respectively. In addition, the loss on ignition of the ash was determined (based on DIN EN ISO 18122), which indicated with a value of 3.41% a good burnout of the solid fuel and low slagging.  (2018) year after Sida establishment calculated on five biological replicates at a planting density of 1, 2, or 4 plants per m 2 . Normal letters indicate the significance between different sowing densities in the individual years of the study. Bold letters indicate the significant differences of the accumulated biomass yield over the years. Values indicated with the same letter are not significantly different (p > .05)

| Sida briquettes
The combustion behavior of Sida briquettes was comparable to the combustion of wood. Overall, the quality of combustion in the stove did not achieve adequate burnout in the gas phase, which in total resulted in very high CO emissions, as presented in Table 1, compared with values of log wood among all average emission values of all tested Sida solid fuels, that is, chips, pellets, and briquettes. To allow a better comparison, Table 1 also represents values from other biogenic solid fuels, that is, Miscanthus pellets, wood pellets, wood chips, and log wood.

| Slagging behavior
None of the Sida fuel trials encountered slagging problems. Thus, Sida differs significantly from other energy plants, such as Miscanthus, which tend to sinter at temperatures of about 1,126°C (calculated), as shown in Figure 3 obtained from the burning of Sida and Miscanthus pellets. Overall, the ashes from the combustion of Sida pellets were relatively homogeneous and fine-grained (Figure 3, left). While only a few smaller agglomerates (<15 mm) could be found in the combustion of Sida pellets, significant slagging was observed in the same furnace under the same conditions when Miscanthus was burnt (agglomerates with diameters of up to 120 mm, Figure 3, right).

| Biomass and ash characterization
Results of the elemental and oxide analysis (Table 2) indicated that Sida is a calcium-silicon-rich to calcium-rich biomass with high amounts of alkali metals, especially potassium. In addition, phosphorous compounds were likely to occur due to significant phosphorus content. With reference to the way of processing the biomass, it could be shown that Sida chips and briquettes were chemically equal, but pellets differed in chemical composition. Sida pellets contained higher amounts of potassium, sulfur, and phosphorus.
With reference to Table 2, the variation of calcium and silica within the Sida chips and briquette samples was very high, also affecting the ash content. Therefore, two exemplary compositions are given for chips and briquettes in Table  2, for each one sample with low SiO 2 and high SiO 2 content, respectively. Since SiO 2 is also enriched in the coarse bottom ash of the respective biomass, a contamination with silica sand is likely. All other elements were stable in amount. The high calcium contents led to very high melting points due to the lack of network formers. The coarse fraction of the bottom ash showed high amounts of unburned carbon. The burnout of the briquettes seemed to be better because of lower carbon content in the coarse fraction.
The variation in chemical composition of fuel and laboratory ash samples was not observed for Sida pellets. The variation between raw material, laboratory ash, and bottom ash was minor, with only small enrichment effects between fine and coarse bottom ash. Nevertheless, Sida pellets also had a high amount of unburned carbon in the coarse fraction. As the oxide compositions indicated, the mineralogy of all samples also varied widely (Table 3). Calcite, hydroxylapatite, and periclase were found in all samples in descending amounts. Also, quartz was detected in every sample, although the characterized chips (#2) and pellet ashes indicated higher amounts than the briquette ash (#1). In contrast to the other ashes, pellets contained a high amount of T A B L E 2 Elemental analysis and oxide analysis of raw Sida chips, briquettes, and pellets, their 550°C laboratory ashes, and different fractions in the bottom ash of the used boiler for combustion analysis (absolute ash content calculated from ICP OES elements as oxides)

Absolute wt% in biomass
Normalized wt% in ash the potassium carbonate (arcanite). Silicates were only found in chips ash. Furthermore, the Sida ashes were rich in phosphorus while the highest content was found in the coarse bottom ash of Sida pellets with 19.5 wt% P 2 O 5 ( Table 2). As the XRD results (Table 3) revealed, phosphorus was mainly bound in calcium phosphates.
All ashes from all tested Sida fuels melted at very high temperatures (Figure 4) due to the high amount of alkaline earth elements. Ash from Sida briquettes showed the highest amount of calcium oxide with about 60 wt%. This sample did not even shrink in the investigated temperature range up to 1,450°C.
From the analysis of the fuels, relevant key figures can be derived that are of great relevance for the assessment of the energy use of Sida biomass. Relevant combustion parameters are presented below and compared with values of other fuels either from own analyses or literature data (Table 4). In addition, a comparison was made with requirements from the standard DIN EN ISO 17225-6, which sets quality requirements for fuel pellets made from non-woody biomasses ( Table 5).

was used to estimate the calorific value (lower heating value [LHV]) from the composition of a fuel (carbon C, hydrogen H, sulphur S, oxygen O):
Further empirical formulas were used to estimate the softening and the flow temperature of the ashes, which are in case of Sida in good agreement with the measured ash fusion behavior.
In Table 4, the results from the elemental analyses and the calculated LHV and melting temperatures of Sida crude biomass (non-processed, i.e., not pelletized or pressed) are compared with other biogenic reference fuels.
Sida is similar to wood in terms of ash softening, and the ash softening temperature higher than that of beech including its bark, poplar, and Miscanthus (Table 4). This shows a significant advantage of Sida over other biogenic energy crops.
The produced Sida pellets were compared in terms of important properties with the requirements for pellets made of non-woody fuels in accordance with DIN EN ISO 17225-6. For the examined parameters, the Sida pellets comply with all the requirements of the standard, with the exception of the content of fines (Table 5).

| Energy yield analysis and calorific evaluation of Sida pellets
The energy yield was analyzed based on the experiment with the Sida pellets due to their stable, representative, and reproducible burning behavior. Within our study, no optimal settings could be found for both the Sida briquettes and Sida chips. Sida pellets allowed, however, very stable experimental conditions. Due to this result, we lay our emphasis on the results obtained from the combustion of Sida pellets in the following. During the experiment, 121.5 kg of Sida pellets were fed into the furnace over a period of approx. 3.8 hr, representing a thermal input of 142.12 kW. A cooling water output of 131 kW was measured. Including the sensible heat introduced via air and fuel, this corresponds to a boiler efficiency of 90.85%.

F I G U R E 4
Ash fusion test of the 550°C laboratory ashes of all Sida fuel types. A decrease in the graph indicates a shrinking or melting event T A B L E 4 Biomass composition of Sida and other biogenic fuels or feedstocks and their calculated calorific value using the Boie formula (given values also extracted from Kaltschmitt et al., 2009), and content of the ash-relevant minerals K, Ca, and Mg in Sida ash compared with other relevant biogenic energy feedstocks

| Life cycle assessment
Depending on the dry biomass and its processing system, different amounts of Sida biomass are necessary for the production of 100 GW heat (FU). The largest input (8,152 kg/ FU) can be seen in process chain SI and is mainly caused by low heating values and relatively large losses of biomass. As shown in Table 6 also other input flows differ for each process chain due to varying parameters (e.g., heating value or water content) and the different design and technologies of the process chains. Water, electricity, and energy resources in general are very low for the production of chips because of reduced technology deployment after harvesting. Fertilizer and diesel inputs correlate with the specific field size. Process chain SII contains extra diesel inputs due to truck-driven conveyor technology for blowing the pellets into the storage by compressed air. The biomass loss represents approximately 9%-12% of the primary biomass, whereby the water loss by drying/evaporation is not listed.
Resulting ash amounts are <4% of the intermediate product quantities.
As a result of normalization, it is possible to identify the five impact categories with the largest PEs, resource depletion, human toxicity, ecotoxicity, and climate change effects. For these effects, the contribution of the various process stages was assessed. Figure 5 indicates that the largest impacts are caused during the biomass production and the combustion phase in general. In principle, this distribution is also valid for the other 11 categories that are not shown. The sum of the 16 shown categories indicates that 33%-36% of the PEs are linked to the biomass production stage, while the combustion stage is responsible for 33%-48% of the PEs in the different scenarios (see lower section of Figure 5). Shares between 9% and 16% can be seen for the ash utilization stage. In contrast to these shares, the impact of the further stages is low (<11%).
In the case of the impact category 'resource depletion, minerals, fossils, and renewables' (RD mfr), the dominating stage in all process chains can be identified as biomass production and its upstream chains (main reason: P-fertilizer production), whereby the production/processing stage also contributes larger shares (main reason: infrastructure/ building). The most clearly expressed share of the 'human toxicity-cancer effect' (HT CE) category can be found in the combustion stage (main reason: furnace production). Further relevant impacts of the HT CE category result from the biomass production stage (main reason: fertilizer production). The 'human toxicity-non-cancer effect' (HT nCE) category is dominated by the ash utilization and the biomass production (main reason: fertilizer production) while the 'ecotoxicity freshwater' (ET Fw) is mainly generated by the ash utilization and combustion stage (main reasons: furnace production). In the 'climate change' category (CC), the combustion stage shows the largest impact in all process chains (main reason: combustion). The stage with the lowest impact can be identified as storage and shipping stage in all categories and process chains. Summarizing it can be stated that the process modules with the largest

T A B L E 6
Major input and output flows of the three process chain variants for the production and use of Sida chips, pellets, and briquettes impacts (PE) can be identified as fertilizer production and use (especially P-and N-fertilizer) during the primary biomass production stage as well as the electricity supply for the manufacturing of the dry biomass. Furthermore, PE increasing process modules can be found as combustion, furnace production, and electricity supply of the combustion stage.
To get a basic impression of the ecological competitiveness, the discussed results of each process chain were compared with other biomass datasets. The aggregated datasets for different biomasses (wood pellets, mixed logs, softwood chips, hardwood chips) show the same magnitude of furnaces' performance classes (25-50 kW) and approximately the same coverage of contained processes. Table 7 shows the F I G U R E 5 Comparison of three process chains by process stages for five selected midpoint impact categories in line with ILCD and summarized values for all 16 ILCD impact categories (B, briquettes; C, chips; CC, climate change; ET Fw, ecotoxicity freshwater; HT CE, human toxicity-cancer effect; HT nCE, human toxicity-non-cancer effect; P, pellets; RD mfr, resource depletion, minerals, fossils, and renewables) T A B L E 7 Comparison of each normalized ILCD impact category for the Sida biomass process chains SI, SII, and SIII with aggregated datasets of alternative biomass energy systems A1-A5 exemplary results for the described process chains as well as five alternative biomass systems (Bauer, 2007). While the results in many categories are in a similar value range (e.g., 'acidification freshwater' AC or 'eutrophication freshwater' EU Fw), it can be stated that several impact categories of the Sida biomass-related process chains have lower values than the alternative systems (e.g., 'Particulate matter/Respiratory inorganics' PM) and vice versa (e.g., RD mfr). Considering the values of all impact categories (added up), the lowest amounts are caused by process chain A3 and the Sida biomass cases, which means that these process chains have the lowest overall impact.

| Biomass production
The planting of Sida using root cuttings demonstrates an effective reproduction method under field conditions (Borkowska & Molas, 2013), among seedlings Šiaudinis et al., 2015), since seed germination was found to be very low (10%, e.g., Borkowska & Wardzińska, 2003), but could be increased significantly by appropriate pre-treatments (Von Gehren & Gansberger, 2017). However, the suppression of weeds is crucial to allow a successful plant establishment.
Research is needed if herbicide or mechanical treatments could be replaced by soil coverage using compostable plastic films or a layer of biomass, for example, mulch or beneficial intercropping using legumes such as white clover, resulting even in an overall biomass increase and nitrogen enrichment (Nabel, Schrey, Temperton, Harrison, & Jablonowski, 2018). This, in return, can also help to reduce evaporation which can be important in years of low precipitation but also bares the risk to compete with Sida for limited resources . The high recorded biomass yields in our present study confirm the major influence of the overall soil quality, precipitation, and local conditions on the Sida plantations compared with other literature data, among fertilization and liming which might be beneficial for plant performance in many cases (Borkowska & Molas, 2013;Borkowska et al., 2009;Šiaudinis, Skuodienė, & Repšienė, 2017). The employed soil compaction after planting of the root cuttings showed no effect on sprouting and plant establishment, underlining the modest nature of Sida as a promising bioenergy crop.

| Economic evaluation
Market estimates for Sida root cuttings are not available since it is not an established energy crop. While Miscanthus rhizomes are sold for approximately 0.16 EUR/piece, Sida root harvest and preparation is more complicated and the supply and demand is lower, we therefore estimate a price of 0.20 EUR/piece used for our calculations, accounting for 2,000, 4,000, or 8,000 EUR/ha, respectively. Depending on the Sida planting density (1, 2, or 4 plants per m 2 ), we estimated costs for initial field preparation and planting accounting for 375, 450, and 600 EUR, respectively. An adequate planting density of perennial energy crops is crucial during the establishment phase, since further modifications of the planting density are not feasible in an existing vegetation. After successful establishment, more than 20 annual harvests are possible (Borkowska & Molas, 2012). Dividing the high establishment costs by 20 harvests, the financial burden can be significantly reduced. In literature, the energy expenses of Sida growing methods (including planting and annual harvesting) reached 8,630-29,264 MJ/ha (Šiaudinis et al., 2015), while our own calculations account for 10,200 MJ ha −1 year −1 , considering the main energy-demanding processes, that is, fertilization and harvest. Annual fertilization and application was estimated using 100 kg N, 100 kg K 2 O, and 120 kg P 2 O 5 /ha (Borkowska et al., 2009), accounting in total for 8,444 MJ/ ha, equaling approx. 280 EUR/ha. Annual harvesting costs were calculated with 317 EUR/ha, and a gasoline consumption of 58.9 L/ha, equaling approx. 2,120 MJ/ha. To date, literature on Sida biomass costs is scarcely available. In Poland, market price estimates for naturally dried Sida chips account for 58.8 EUR/t (Stolarski et al., 2013). Such estimates are difficult for Germany or elsewhere in Europe due to missing Sida production on a comparable level. However, there is a market for Miscanthus biomass accounting for 80-100 EUR/t, which can be considered as a price making the cultivation and harvest of such perennial energy crops feasible. However, the applied calculation method can be questioned due to the variable humidity of the biomass, suggesting a payment for the energy contained in the biomass in EUR/GJ.
Employing higher plant densities, that is, 2 or 4 plants versus 1 plant per m 2 , resulted in higher biomass overall. The biomass yield increased by approx. 30% and 60% employing 2 or 4 plants, respectively, compared with 1 plant per m 2 . To obtain the first 30% yield increase (i.e., 2 plants per m 2 ), the investment cost for field establishment almost doubled; however, to gain the next 30% (i.e., 4 plants per m 2 ), the establishment costs increase exponentially, favoring 2 plants per m 2 overall. Additionally, higher planting densities potentially increase infections with fungi like Sclerotinia sclerotiorum, resulting in additional costs for fungicides (Remlein-Starosta, Krzymińska, Kowalska, & Bocianowski, 2016). All mentioned expense factors result in an estimated market net price of 80-100 EUR/t for naturally dried chips of Sida. This price is up to 70% higher compared with earlier data from literature accounting for 58.8 EUR/t (Stolarski et al., 2013), mirroring the different expenses locally for resources and establishment costs allowing an economically viable cultivation and harvest of Sida as a sustainable biomass feedstock for energy generation.

| Biomass processing and energy yield
Dried Sida biomass can easily be harvested using conventional crop choppers as applied for maize. The further treatment of biomass by pelleting offers the chance to change and standardize the physical properties of the Sida biomass. Compaction achieves not only a higher bulk density and thus a lower specific transport weight and a higher energy density but also improves the combustion properties by reducing the fine particle content and homogenizing the particle size distribution, simplifying significantly the transport and handling of the fuel. However, the bulk density of Sida pellets was lower compared with previous studies (662 kg/m 3 vs. 969 kg/ m 3 ; Šiaudinis et al., 2015), while still meeting the requirements of the norm DIN EN ISO 17225-6 for pellets made of non-woody biomass. The power consumption needed for the Sida pellet production (99.4 kWh/Mg) was in the usual range on pilot plant scale (e.g., 89.2 kWh/Mg wood pellets), but well above the energy requirements of large-scale pelletizing plants, which lie in a range of 40 kWh/Mg pellets, taking only the pressing itself into account (Hartmann, Reisinger, Turowski, & Roßmann, 2013;Hasler & Nussbaumer, 2001). According to DIN EN ISO 17225-6, which is applicable for non-woody biomass fuels, the content of fines (<3.15 mm) is limited to <1%. The norm distinguishes between quality A (fines < 2%) and B (fines < 3%). The determined fines content of 3.8% is due to the production on pilot plant scale. In large-scale plants, an automated recirculation of the fine content takes place via pellet sieving. The increased fine content has to be taken into account when assessing the results from combustion tests. However, all further requirements for pellets made of non-woody fuels were in accordance with DIN EN ISO 17225-6.

| Combustion and emission characteristics
From all tested Sida fuels, pellets showed the best processing and burnout behavior. In contrast, shredded Sida (i.e., chips) cannot be used without technological adaptation in the investigated type of furnace, such as adjustment of the conveyor system to allow a constant material flow into the combustion chamber. In addition, it could be observed that finer particles were captured by primary air due to the low specific gravity and transported to the end of the grate. This observation is also evident in the determination of the loss on ignition of the ash obtained from Sida chips combustion, which shows an incomplete solid burnout with 16.91%. Furthermore, the residual moisture content in the Sida chips accounting for approx. 19% at harvest may have contributed to the observed initial drop in temperature and higher CO emissions for Sida chips. Interestingly, the combustion behavior and emission values of Sida briquettes were comparable to log wood, showing that the compaction of Sida biomass seems to be advisable for combustion purposes, as well as for transport and storage, compared with the loose Sida chips. As shown, our calculated and measured LHV of Sida was in accordance with the mean value reported in literature (Nahm & Morhart, 2018), and was comparable to other biogenic sources as indicated in this study. The elemental composition of the evaluated solid Sida fuels was in accordance with literature data as well (Stolarski et al., 2014).
Since Sida pellets demonstrated best combustion behavior overall, it is noteworthy that the number of firing systems for pellets increased linearly during the last 9 years in Germany, accounting for almost half a million installations in 2019 (FNR, 2019). While private households are by 22.2% the second largest user group of solid fuels from woody resources (FNR, 2019), Sida pellets may substitute and contribute to the supply of such biogenic fuels for heat and energy generation.
Even though the content of sulfur as a highly eroding element was low in Sida biomass itself, meeting the DIN EN ISO 17225-6 requirements, the SO 2 emissions were relatively high for all solid Sida fuels compared with Miscanthus pellets, wood pellets or chips, except for log wood. Interestingly, our results on SO 2 were the exception, and cannot be explained with the S content in the biomass or the added molasses for the pellet production. Though unlikely, residual rapeseed cake in the pellets production chain may have contributed to the observed high SO 2 and NO x emissions. Our values for NO x and CO in the exhaust gases were significantly smaller when compared to literature data on the combustion of Sida pellets (Zajac, Szyszlak-Barglowicz, Slowik, Wasilewski, & Kuranc, 2017). However, the high values given in the reference were explained by technical problems and incomplete pellet combustion. High fluctuation in magnitude of NO, NO x , and CO depends on fuel feeding methods, the natural N content of the biomass, the combustion temperature, and oxygen flow (Juszczak, 2014). High CO values determined for Sida briquettes may be related to the solid, cylindrical nature of the tested briquettes, which likely resulted in incomplete burnout and therefore higher CO evolution.

| Ash analysis
A lower ash content in biomass used for combustion purposes implicates a higher energy density, and is desirable to avoid high dust loads and therefore pollution and waste. The ash content for all Sida solid fuels was lower as or equal to values reported in other studies on Sida as a solid fuel (2.1%-4.7% vs. 6% or 2.8%, respectively; Šiaudinis et al., 2015;Stolarski et al., 2013), and generally in the lower range when compared to a vast number of biogenic fuels (Vassilev et al., 2017). Sida ash values might be related to a generally high uncertainty factor in the analytical methodology and inhomogeneity of the Sida biomass composition from different locations. However, an ash content of approx. 3% for all tested Sida fuels, and particularly for the pellets meets the DIN EN ISO 17225-6 requirements accounting for a value of <6%. As reported in our earlier study , we here report high ash melting temperatures which are in line with a previous investigation on Sida used as a solid fuel (von Gehren et al., 2019). The high melting temperatures were due to the lack of network formers like silica or alumina species. Generally, those ashes are not critical for slagging, even though those high alkaline contents can lead to severe sintering. Concerning the network theory (Vorres, 1987), an even combination of alkaline (earth) metals and aluminosilicates including iron and phosphorus should be avoided because of low sintering temperatures and subsequent sintering or slagging on the combustion chamber walls and superheater, or reheater surfaces, respectively, which can increase corrosion and causes reduced heat transfer and thus lower efficiency of the furnace.
The high amount of silicates found in the chips ashes might be due to missing pre-processing in comparison to the other ashes preserving more clays from the Sida experimental field.
The Sida ashes were rich in phosphorus (P), which is an important element if the ash shall be recycled as fertilizer. As shown, phosphorus was mainly bound in calcium phosphates which are only little soluble in water and therefore relatively immobile. Therefore, the ash could likely act as long-term fertilizer, which is a promising option presented in a previous study on other biogenic ashes with similar P contents (Schiemenz & Eichler-Löbermann, 2010). However, it needs further investigation if such high P contents in Sida ashes are generally found elsewhere and if the P-forms are plant available.
Irrespective of the vast variation in the elemental composition of all investigated Sida fuel ashes, the ash melting point and most detected elements were in a similar range as reported elsewhere (Kowalczyk-Jusko, 2009). The high variation in ash composition has to be linked to the different Sida fuels tested in our study, irrespective the fact that the analytical methods display a variation itself even though sample preparation was conducted accurately.

| Life cycle assessment
The presented results show advantages and disadvantages of the three process chains considered. It can be stated that on the one hand process chain SII (pellets) needs the lowest biomass input and process chain SIII (briquettes) also tends into this direction in comparison with process chain SI (chips). On the other hand, the specific water, direct electricity, and energy resources consumption are significantly higher in these process chains. Adding the output category 'Sida biomass loss' and 'ash', SI and SII are on a par. Overall, the results show the general effects of an advanced processing after shredding.
By separating each process chain into process stages, possible savings of input flows that have the largest effects can identify starting points for optimization. Involved parties (grower, biomass processors, or heat producer) are able to identify possibilities to influence the ecological impacts. Biomass production and combustion have been identified to be mainly responsible for most impacts, and are therefore suitable intervention points for further analysis.
The comparison of LCIA results among Sida biomass process chains themselves can play a big role in identifying the most ecologically favorable product. The present results show the lowest ecological impact for Sida pellets, closely followed by Sida briquettes and chips. A definite ranking would need a clearer differentiation within the results. The comparison with alternative technologies generates arguments for the competitiveness of Sida biomass. Causes for environmental impacts must be divided into influenceable and non-influenceable for the producer. While fertilizer production itself is not modifiable for consumers, the using conditions (e.g., mass, source, etc.) are. The reduction of fertilizers, the use of biogenic residues, and the choice of location are starting points for optimization of ecological effects, as shown in two earlier studies on the establishment of Sida under marginal soil conditions (Nabel, Schrey, Poorter, et al., 2018;. The furnace production takes place outside the area of influence for the heat producers. To get better values for the combustion process, the optimization of the energy carrier's combustion behavior is an option. Nevertheless, it was illustrated that the use of Sida biomass led to significantly lower PE values than the most alternative biomass routes with wood chips or logs, etc. (four of five cases, Table 7). In the clearest case, an advantage of nearly 50% of the absolute PE value was determined for Sida biomass. Furthermore, a comparison with suitable datasets on fossil energy generation as well as other alternative biomasses (e.g., Miscanthus) should be realized. In the present study, this was not possible due to excessive deviations of performance classes (furnace) and other non-compliant conditions. The indicated advantageousness of Sida biomass can be underpinned by further research on >15 different datasets (unpublished data), where >70% of considered alternative process chains turn out to have a larger impact than Sida. Further studies should include the currently much debated topics of soil organic carbon and CO 2 sequestration and, where appropriate, be based on new corresponding studies. From today's point of view, the ecological impact of Sida biomass as an energy carrier makes it a useful alternative to the most compared established solid fuels.